原创

纲要

Neural Networks(神经网络) and Deep Learning

will learn how to build a neural network and how to train them on data

Improving Deep Neural Networks: Hyperparameter tuning(超参数调整)，Regularization(正则化) and Optimization(高级优化算法)

Structuring your Machine Learning project

Convolutional Neural Networks(卷积神经网络，CNN)

always apply for images

National Language Procession: Building sequence models

include Recurrent Neural Networks(循环神经网络，RNN)

1.2 什么是神经学习

deep learning refers to training neural networks

image-20230628162111118

• standard neural network

  • Home feature->Price

  • Ad,user info->Click on ad?(0/1)

• CNN

  • image->object

• RNN

  • sequence data
    • andio
    • language

神经网络适用于两种数据

• 结构化数据
  • 数据的数据库
• 非结构化数据
  • audio or image

为什么深度学习会兴起

introduction answer:

CD(wrong):

1. What does the analogy “AI is the new electricity” refer to?(和“AI 是新电力”相类似的说法是

什么？)

【 】AI is powering personal devices in our homes and offices, similar to electricity.(AI 为我们

的家庭和办公室的个人设备供电，类似于电力。)

【 】Through the “smart grid”, AI is delivering a new wave of electricity.(通过“智能电网”，AI

提供新的电能。)

【 】AI runs on computers and is thus powered by electricity, but it is letting computers do

things not possible before.(AI 在计算机上运行，并由电力驱动，但是它正在让以前的计算机不

能做的事情变为可能。)

【★】Similar to electricity starting about 100 years ago, AI is transforming multiple

industries.(就像 100 年前产生电能一样，AI 正在改变很多的行业。)

​ AC ACD F C F F

C(not full):

8. Why is an RNN (Recurrent Neural Network) used for machine translation, say translating

English to French? (Check all that apply.)(为什么在上 RNN（循环神经网络）可以应用机器翻

译将英语翻译成法语？)

【★】It can be trained as a supervised learning problem. (因为它可以被用做监督学习。)

【 】It is strictly more powerful than a Convolutional Neural Network (CNN).(严格意义上它比

卷积神经网络（CNN）效果更好。)

【★】It is applicable when the input/output is a sequence (e.g., a sequence of words). (它比较

适合用于当输入/输出是一个序列的时候（例如：一个单词序列）)

【 】RNNs represent the recurrent process of Idea->Code->Experiment->Idea->….(RNNs 代表

递归过程：想法->编码->实验->想法->…)

​ AB